source: heimdal/trunk/lib/wind/utf8.c

/*
 * Copyright (c) 2004, 2006, 2007, 2008 Kungliga Tekniska Högskolan
 * (Royal Institute of Technology, Stockholm, Sweden).
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 *
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 *
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * 3. Neither the name of the Institute nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE INSTITUTE AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE INSTITUTE OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 */

#include <config.h>
#include "windlocl.h"

static int
utf8toutf32(const unsigned char **pp, uint32_t *out)
{
    const unsigned char *p = *pp;
    unsigned c = *p;

    if (c & 0x80) {
        if ((c & 0xE0) == 0xC0) {
            const unsigned c2 = *++p;
            if ((c2 & 0xC0) == 0x80) {
                *out =  ((c  & 0x1F) << 6)
                    | (c2 & 0x3F);
            } else {
                return WIND_ERR_INVALID_UTF8;
            }
        } else if ((c & 0xF0) == 0xE0) {
            const unsigned c2 = *++p;
            if ((c2 & 0xC0) == 0x80) {
                const unsigned c3 = *++p;
                if ((c3 & 0xC0) == 0x80) {
                    *out =   ((c  & 0x0F) << 12)
                        | ((c2 & 0x3F) << 6)
                        |  (c3 & 0x3F);
                } else {
                    return WIND_ERR_INVALID_UTF8;
                }
            } else {
                return WIND_ERR_INVALID_UTF8;
            }
        } else if ((c & 0xF8) == 0xF0) {
            const unsigned c2 = *++p;
            if ((c2 & 0xC0) == 0x80) {
                const unsigned c3 = *++p;
                if ((c3 & 0xC0) == 0x80) {
                    const unsigned c4 = *++p;
                    if ((c4 & 0xC0) == 0x80) {
                        *out =   ((c  & 0x07) << 18)
                            | ((c2 & 0x3F) << 12)
                            | ((c3 & 0x3F) <<  6)
                            |  (c4 & 0x3F);
                    } else {
                        return WIND_ERR_INVALID_UTF8;
                    }
                } else {
                    return WIND_ERR_INVALID_UTF8;
                }
            } else {
                return WIND_ERR_INVALID_UTF8;
            }
        } else {
            return WIND_ERR_INVALID_UTF8;
        }
    } else {
        *out = c;
    }

    *pp = p;

    return 0;
}

/**
 * Convert an UTF-8 string to an UCS4 string.
 *
 * @param in an UTF-8 string to convert.
 * @param out the resulting UCS4 strint, must be at least
 * wind_utf8ucs4_length() long.  If out is NULL, the function will
 * calculate the needed space for the out variable (just like
 * wind_utf8ucs4_length()).
 * @param out_len before processing out_len should be the length of
 * the out variable, after processing it will be the length of the out
 * string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_utf8ucs4(const char *in, uint32_t *out, size_t *out_len)
{
    const unsigned char *p;
    size_t o = 0;
    int ret;

    for (p = (const unsigned char *)in; *p != '\0'; ++p) {
        uint32_t u;

        ret = utf8toutf32(&p, &u);
        if (ret)
            return ret;

        if (out) {
            if (o >= *out_len)
                return WIND_ERR_OVERRUN;
            out[o] = u;
        }
        o++;
    }
    *out_len = o;
    return 0;
}

/**
 * Calculate the length of from converting a UTF-8 string to a UCS4
 * string.
 *
 * @param in an UTF-8 string to convert.
 * @param out_len the length of the resulting UCS4 string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_utf8ucs4_length(const char *in, size_t *out_len)
{
    return wind_utf8ucs4(in, NULL, out_len);
}

static const char first_char[4] =
    { 0x00, 0xC0, 0xE0, 0xF0 };

/**
 * Convert an UCS4 string to a UTF-8 string.
 *
 * @param in an UCS4 string to convert.
 * @param in_len the length input array.

 * @param out the resulting UTF-8 strint, must be at least
 * wind_ucs4utf8_length() + 1 long (the extra char for the NUL).  If
 * out is NULL, the function will calculate the needed space for the
 * out variable (just like wind_ucs4utf8_length()).

 * @param out_len before processing out_len should be the length of
 * the out variable, after processing it will be the length of the out
 * string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_ucs4utf8(const uint32_t *in, size_t in_len, char *out, size_t *out_len)
{
    uint32_t ch;
    size_t i, len, o;

    for (o = 0, i = 0; i < in_len; i++) {
        ch = in[i];

        if (ch < 0x80) {
            len = 1;
        } else if (ch < 0x800) {
            len = 2;
        } else if (ch < 0x10000) {
            len = 3;
        } else if (ch <= 0x10FFFF) {
            len = 4;
        } else
            return WIND_ERR_INVALID_UTF32;

        o += len;

        if (out) {
            if (o >= *out_len)
                return WIND_ERR_OVERRUN;

            switch(len) {
            case 4:
                out[3] = (ch | 0x80) & 0xbf;
                ch = ch >> 6;
            case 3:
                out[2] = (ch | 0x80) & 0xbf;
                ch = ch >> 6;
            case 2:
                out[1] = (ch | 0x80) & 0xbf;
                ch = ch >> 6;
            case 1:
                out[0] = ch | first_char[len - 1];
            }
        }
        out += len;
    }
    if (out) {
        if (o + 1 >= *out_len)
            return WIND_ERR_OVERRUN;
        *out = '\0';
    }
    *out_len = o;
    return 0;
}

/**
 * Calculate the length of from converting a UCS4 string to an UTF-8 string.
 *
 * @param in an UCS4 string to convert.
 * @param in_len the length of UCS4 string to convert.
 * @param out_len the length of the resulting UTF-8 string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_ucs4utf8_length(const uint32_t *in, size_t in_len, size_t *out_len)
{
    return wind_ucs4utf8(in, in_len, NULL, out_len);
}

/**
 * Read in an UCS2 from a buffer.
 *
 * @param ptr The input buffer to read from.
 * @param len the length of the input buffer.
 * @param flags Flags to control the behavior of the function.
 * @param out the output UCS2, the array must be at least out/2 long.
 * @param out_len the output length
 *
 * @return returns 0 on success, an wind error code otherwise.
 * @ingroup wind
 */

int
wind_ucs2read(const void *ptr, size_t len, unsigned int *flags,
              uint16_t *out, size_t *out_len)
{
    const unsigned char *p = ptr;
    int little = ((*flags) & WIND_RW_LE);
    size_t olen = *out_len;

    /** if len is zero, flags are unchanged */
    if (len == 0) {
        *out_len = 0;
        return 0;
    }

    /** if len is odd, WIND_ERR_LENGTH_NOT_MOD2 is returned */
    if (len & 1)
        return WIND_ERR_LENGTH_NOT_MOD2;

    /**
     * If the flags WIND_RW_BOM is set, check for BOM. If not BOM is
     * found, check is LE/BE flag is already and use that otherwise
     * fail with WIND_ERR_NO_BOM. When done, clear WIND_RW_BOM and
     * the LE/BE flag and set the resulting LE/BE flag.
     */
    if ((*flags) & WIND_RW_BOM) {
        uint16_t bom = (p[0] << 8) + p[1];
        if (bom == 0xfffe || bom == 0xfeff) {
            little = (bom == 0xfffe);
            p += 2;
            len -= 2;
        } else if (((*flags) & (WIND_RW_LE|WIND_RW_BE)) != 0) {
            /* little already set */
        } else
            return WIND_ERR_NO_BOM;
        *flags = ((*flags) & ~(WIND_RW_BOM|WIND_RW_LE|WIND_RW_BE));
        *flags |= little ? WIND_RW_LE : WIND_RW_BE;
    }

    while (len) {
        if (olen < 1)
            return WIND_ERR_OVERRUN;
        if (little)
            *out = (p[1] << 8) + p[0];
        else
            *out = (p[0] << 8) + p[1];
        out++; p += 2; len -= 2; olen--;
    }
    *out_len -= olen;
    return 0;
}

/**
 * Write an UCS2 string to a buffer.
 *
 * @param in The input UCS2 string.
 * @param in_len the length of the input buffer.
 * @param flags Flags to control the behavior of the function.
 * @param ptr The input buffer to write to, the array must be at least
 * (in + 1) * 2 bytes long.
 * @param out_len the output length
 *
 * @return returns 0 on success, an wind error code otherwise.
 * @ingroup wind
 */

int
wind_ucs2write(const uint16_t *in, size_t in_len, unsigned int *flags,
               void *ptr, size_t *out_len)
{
    unsigned char *p = ptr;
    size_t len = *out_len;

    /** If in buffer is not of length be mod 2, WIND_ERR_LENGTH_NOT_MOD2 is returned*/
    if (len & 1)
        return WIND_ERR_LENGTH_NOT_MOD2;

    /** On zero input length, flags are preserved */
    if (in_len == 0) {
        *out_len = 0;
        return 0;
    }
    /** If flags have WIND_RW_BOM set, the byte order mark is written
     * first to the output data */
    if ((*flags) & WIND_RW_BOM) {
        uint16_t bom = 0xfffe;

        if (len < 2)
            return WIND_ERR_OVERRUN;

        if ((*flags) & WIND_RW_LE) {
            p[0] = (bom     ) & 0xff;
            p[1] = (bom >> 8) & 0xff;
        } else {
            p[1] = (bom     ) & 0xff;
            p[0] = (bom >> 8) & 0xff;
        }
        len -= 2;
    }

    while (in_len) {
        /** If the output wont fit into out_len, WIND_ERR_OVERRUN is returned */
        if (len < 2)
            return WIND_ERR_OVERRUN;
        if ((*flags) & WIND_RW_LE) {
            p[0] = (in[0]     ) & 0xff;
            p[1] = (in[0] >> 8) & 0xff;
        } else {
            p[1] = (in[0]     ) & 0xff;
            p[0] = (in[0] >> 8) & 0xff;
        }
        len -= 2;
        in_len--;
        p += 2;
        in++;
    }
    *out_len -= len;
    return 0;
}


/**
 * Convert an UTF-8 string to an UCS2 string.
 *
 * @param in an UTF-8 string to convert.
 * @param out the resulting UCS2 strint, must be at least
 * wind_utf8ucs2_length() long.  If out is NULL, the function will
 * calculate the needed space for the out variable (just like
 * wind_utf8ucs2_length()).
 * @param out_len before processing out_len should be the length of
 * the out variable, after processing it will be the length of the out
 * string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_utf8ucs2(const char *in, uint16_t *out, size_t *out_len)
{
    const unsigned char *p;
    size_t o = 0;
    int ret;

    for (p = (const unsigned char *)in; *p != '\0'; ++p) {
        uint32_t u;

        ret = utf8toutf32(&p, &u);
        if (ret)
            return ret;

        if (u & 0xffff0000)
            return WIND_ERR_NOT_UTF16;

        if (out) {
            if (o >= *out_len)
                return WIND_ERR_OVERRUN;
            out[o] = u;
        }
        o++;
    }
    *out_len = o;
    return 0;
}

/**
 * Calculate the length of from converting a UTF-8 string to a UCS2
 * string.
 *
 * @param in an UTF-8 string to convert.
 * @param out_len the length of the resulting UCS4 string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_utf8ucs2_length(const char *in, size_t *out_len)
{
    return wind_utf8ucs2(in, NULL, out_len);
}

/**
 * Convert an UCS2 string to a UTF-8 string.
 *
 * @param in an UCS2 string to convert.
 * @param in_len the length of the in UCS2 string.
 * @param out the resulting UTF-8 strint, must be at least
 * wind_ucs2utf8_length() long.  If out is NULL, the function will
 * calculate the needed space for the out variable (just like
 * wind_ucs2utf8_length()).
 * @param out_len before processing out_len should be the length of
 * the out variable, after processing it will be the length of the out
 * string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_ucs2utf8(const uint16_t *in, size_t in_len, char *out, size_t *out_len)
{
    uint16_t ch;
    size_t i, len, o;

    for (o = 0, i = 0; i < in_len; i++) {
        ch = in[i];

        if (ch < 0x80) {
            len = 1;
        } else if (ch < 0x800) {
            len = 2;
        } else
            len = 3;

        o += len;

        if (out) {
            if (o >= *out_len)
                return WIND_ERR_OVERRUN;

            switch(len) {
            case 3:
                out[2] = (ch | 0x80) & 0xbf;
                ch = ch >> 6;
            case 2:
                out[1] = (ch | 0x80) & 0xbf;
                ch = ch >> 6;
            case 1:
                out[0] = ch | first_char[len - 1];
            }
            out += len;
        }
    }
    if (out) {
        if (o >= *out_len)
            return WIND_ERR_OVERRUN;
        *out = '\0';
    }
    *out_len = o;
    return 0;
}

/**
 * Calculate the length of from converting a UCS2 string to an UTF-8 string.
 *
 * @param in an UCS2 string to convert.
 * @param in_len an UCS2 string length to convert.
 * @param out_len the length of the resulting UTF-8 string.
 *
 * @return returns 0 on success, an wind error code otherwise
 * @ingroup wind
 */

int
wind_ucs2utf8_length(const uint16_t *in, size_t in_len, size_t *out_len)
{
    return wind_ucs2utf8(in, in_len, NULL, out_len);
}